Environmental Regulation vs. Perceived Value About Manure and Sewage Resource Utilization in Chinese Dairy Farms

Abstract

Clarifying the key driving factors behind the adoption of manure resource utilization technology and promoting its widespread application are crucial for achieving the high-quality development of animal husbandry. This study analyses survey data from 412 large-scale dairy farms across 23 provinces in China. The Cov-AHP method is used to measure the adoption intensity of technology and analyse its mechanisms and underlying logic. The results indicate that value perception, particularly economic value perception, is the strongest driver of adoption intensity. Although the direct effect of environmental regulation is limited, it significantly amplifies the influence of value perception—particularly economic value perception—on technology adoption intensity, especially in large-scale farms. Furthermore, incentive-based regulations (e.g., government subsidies) markedly promote the adoption of manure resource utilization technologies, whereas constraint-based measures (e.g., fines) exert stronger effects on small-scale farms. Additionally, demonstration farms serve as critical catalysts for disseminating best practices and accelerating technology adoption. This study suggests that policies should integrate value perception with targeted financial subsidies and regulatory measures to improve technology adoption, especially with support for small-scale farms. By leveraging demonstration farms to promote successful experiences, the comprehensive adoption of manure resource utilization technologies across the industry can be further improved.

1. Introduction

Results from China’s second national census on pollution sources indicate that the comprehensive utilization rate of livestock and poultry manure reached 78% by the end of 2022 [1,2,3,4]. The “Action Plan for Agricultural and Rural Pollution Control (2021–2025)” issued by the Ministry of Ecology and Environment of China aims to further increase this rate to over 80% by 2025 [5]. Utilizing livestock manure and sewage resources (MSRs) not only minimizes discharge but also transforms waste into productive resources, thereby enhancing the economic viability of dairy farming and increasing farmers’ income [6]. Optimizing the adoption strategy of MSR utilization technology is essential for achieving agricultural resource conservation and promoting green and low-carbon development. While significant progress has been made in utilizing MSRs, with end-of-pipe treatment facilities equipped in nearly all scale dairy farms, in 2023 the Ministry of Agriculture and Rural Affairs (MARA) reported that China had approximately 15,000 large-scale dairy farms (herd size >100 head) [7], among which a considerable number of farms still inadequately manage MSRs at their source and during the mid-process stages. Additionally, source reduction includes the application of environmentally friendly feeds, rainwater–sludge separation, dry manure measures, automatic spraying, sewage recovery and recycling, etc. Process control includes the use of composting and fermentation agents, deodorants, and bacterial agents for faecal water treatment [7]. Addressing these challenges is fundamental for enhancing the effectiveness of MSR utilization.

Scholars have extensively studied the factors influencing the technology adoption of MSR utilization. Farms, as primary adopters, are influenced by internal factors such as individual characteristics, farm resources, and behavioural attitudes [8,9,10,11]. On the other hand, MSR utilization presents an externality issue, where external factors such as government policies, subsidies, supervision, and agricultural extension services also shape farmer behaviour [12]. However, the existing research has not reached a consensus on the direction and extent of these influences. While some studies suggest that government subsidies can encourage green practices and increase MSR utilization [13,14,15], others argue that the effectiveness of policies such as environmental regulations in promoting MSR utilization and reducing agricultural pollution requires further validation. Wang et al. (2023) propose that while subsidies for equipment purchases or land rentals are helpful [16,17], they have limited impact on MSR utilization unless combined with penalties or taxes on non-compliant manure and sewage discharge to better internalize production externalities.

In addition, the previous research on the adoption of MSR utilization technologies has primarily focused on the impact mechanisms of value perception or environmental regulation on the decision to adopt single end-of-pipe treatment technologies. This approach neglects the interconnections between technologies, overlooking the fact that MSR utilization is a combination of a series of technologies, including source reduction, process control, and end-of-pipe treatment [18,19]. Although a small number of scholars have examined the comprehensive process of MSR utilization technologies, the assessment of these technologies has predominantly relied on subjective scoring, which lacks objectivity [15,20]. Therefore, it is essential to adopt a scientific approach to measure the comprehensiveness of MSR utilization technology applications. This will allow us to explore whether the enhancement of MSR utilization technology adoption in farms is primarily driven by intrinsic factors, such as value perception, or by external factors, such as environmental regulation, or whether it results from a combined influence of both. Further investigation into these mechanisms is warranted.

This study is based on field survey data from 412 scale dairy farms across 23 provinces in China. Using the Cov-AHP method, it calculates the adoption intensity index of MSR utilization technology, incorporating source reduction, process control, and end-of-pipe treatment technologies. The study constructs a theoretical analysis framework for farm technology adoption decision making, systematically examining the intrinsic motivations and external drivers influencing value perceptions of adopting MSR utilization technology. It explores their impact on enhancing adoption intensity, analyses their underlying mechanisms and internal logic, and provides empirical evidence and guidelines for standardizing MSR utilization behaviour, improving technical systems, and optimizing related policies. The main contributions of this study are twofold. Firstly, in terms of research content, it comprehensively explores the mechanisms influencing technology adoption of MSR utilization in scale dairy farms, integrating internal and external factors to overcome the biases of previous single-perspective studies. Secondly, in terms of methodology, it scientifically measures the adoption intensity of integrated technologies for MSR utilization using an enhanced hierarchical analysis method. This approach considers the technical aspects of source reduction, process control, and end-of-pipe treatment, thereby making the variables more representative and credible.

2. Theoretical Structure and Research Hypotheses

The Theory of Planned Behaviour states that farm behavioural decisions hinge on attitudes shaped by value perception. According to the Theory of Perceived Value [21], adopting MSR utilization technology on farms results from weighing perceived benefits against risks. Motivation theory suggests that farm technology adoption is influenced by both internal factors and external policies [22]. Government policies act as substitutes and supplements in the market economy, promoting technology adoption through regulations and incentives. This dual role not only boosts adoption intensity but also supports the sustainability of MSR utilization practices on farms.

Value perception and the adoption of MSR utilization technology

Farms’ economic choices result from their resource endowment, individual cognition, and external conditions, resulting in what is termed “emotional choice” [23,24]. This process involves evaluating whether economic behaviours meet their expectations, including perceptions of economic, social, and ecological values. Value perception, as defined by perceived value theory [25], emphasizes that enhancing individuals’ understanding of the economic benefits and risks associated with certain behaviours promotes decision making. MSR utilization technology, with its significant externalities aimed at achieving green production and environmental improvements, can see increased adoption rates through improved perceptions of its ecological, environmental, and social impacts [25,26,27]. Perceived Value Theory posits that individuals make decisions based on perceived benefits and risks. In the case of dairy farms, economic value perceptions (such as profitability and cost-saving) are typically more influential on larger farms, which have the financial capacity to absorb new technologies. In contrast, ecological value perception (such as reducing pollution) might be more relevant for smaller or regionally less developed farms, where environmental sustainability can offer competitive advantages or align with local community norms.

Based on this, the following hypothesis is proposed:

H1. 

Value perceptions on economic, social, and ecological factors significantly influence the adoption intensity of MSR utilization technology on farms, with economic value perception having a stronger effect on larger farms, while ecological perceptions are more influential for small farms.

Environmental regulation and the adoption of MSR utilization technology

Environmental issues are typical externalities and the implementation of manure utilization on dairy farms exemplifies this. According to Environmental Economics Theory, regulations can internalize externalities by encouraging farms to adopt more sustainable practices. However, their effectiveness is likely to vary based on farm-specific characteristics. For example, incentive-based regulations (such as subsidies or tax breaks) are typically more attractive to larger farms with greater financial flexibility. On the other hand, constraint-based regulations (such as fines or penalties for non-compliance) may have a stronger impact on smaller farms, which often have limited capital and are more vulnerable to financial penalties [28,29]. Based on this, the following hypothesis is proposed:

H2. 

Environmental regulations have significant scale heterogeneity in promoting the adoption of MSR utilization technology, with incentive-based regulations being more effective for large-scale farms, while constraint-based regulations have a greater impact on small-scale farms.

Moderating effect of environmental regulation on value perception and adoption of MSR utilization technology

Economic activities entail making rational decisions aimed at maximizing interests within defined institutional frameworks. In studying the adoption behaviour of pig manure utilization across multiple provinces, Fei et al. (2023) found that perceived value directly motivates farmers to adopt this technology [30]. However, Yu et al. (2019) noted that while social value perception has a minimal impact on adoption, environmental regulations significantly influence farmers’ choices [31]. In fact, utilizing MSRs involves a complex system requiring new technologies and equipment to boost productivity, alongside promoting effective production models to convert potential into tangible productivity [32]. Enhancing farms’ value perception plays a crucial role in fostering internal motivation, while robust policy frameworks offer external support. Large-scale dairy farms place a significantly higher emphasis on cost control than smaller farms due to their enormous scale and substantial upfront investments, which create high fixed costs, coupled with intense market competition pressure. Consequently, large dairy farms highly emphasize cost awareness, efficiency optimization, and meticulous management in every aspect of their operations. Incentive-based policies (e.g., subsidies, tax breaks) can make the perceived benefits more tangible, directly enhancing the adoption of resource-efficient technologies, especially in farms where cost considerations are more pronounced (usually larger farms). This moderation effect is expected to be less prominent for smaller farms, which might face greater constraints on implementing technology due to financial limitations.

The impact of environmental regulations on farm economic valuation exhibits scale heterogeneity. Economic incentive policies (e.g., subsidies, tax credits) more effectively enhance perceived economic value in large-scale farms due to their economies of scale and resource integration capabilities, which efficiently translate policy benefits into economic gains [33]. Conversely, constraint-based regulations (e.g., pollution standards, non-compliance penalties) disproportionately affect small-to-medium farms, given their higher compliance cost burden, limited technological adaptability, and the direct operational pressure from environmental controls (e.g., waste treatment) that may breach survival cost thresholds. Based on this, the following hypothesis is proposed:

H3. 

Environmental regulations, particularly incentive-type regulations, moderate the relationship between economic value perception and technology adoption intensity for dairy farms, with stronger effects in larger farms.

Based on the theoretical analysis presented above, a theoretical framework is constructed illustrating the interplay of value perception, environmental regulation, and the adoption decision of MSR utilization technology (Figure 1).

3. Materials and Methods

3.1. Materials

The data for this article comes from a survey conducted by the Institute of Agricultural Information of the Chinese Academy of Agricultural Sciences, covering 420 scale dairy farms across 23 provinces in China from March to June 2023. The survey included five major dairy production areas: North China (Henan, Hebei, Shandong, and Shaanxi), Northeast China and Inner Mongolia (Inner Mongolia, Heilongjiang, and Liaoning), Northwest China (Shaanxi, Ningxia, Gansu, and Xinjiang), Southern China (Anhui, Fujian, Guangdong, Yunnan, Guizhou, Sichuan, Hunan, and Jiangsu), and areas around large cities (Beijing, Tianjin, Shanghai, and Chongqing). A combined online and offline survey methodology was employed. Prior to field surveys, the research team underwent multiple training sessions to ensure thorough familiarity with the questionnaire. Offline surveys were conducted via face-to-face interviews, utilizing colloquial language for questioning and recording responses to ensure data authenticity and reliability. Online surveys were distributed using the Questionnaire Star platform, with a one-month collection window. The questionnaire comprehensively covered four key domains: (1) Basic Information, including farm operator details, fundamental farm characteristics, and employment status; (2) Feed and Nutrition, encompassing the adoption of environmentally friendly feed (including year of adoption), willingness to adopt such feed, and the influence of government promotion; (3) Manure Treatment and Utilization, covering manure removal methods, wastewater treatment methods, manure utilization methods, willingness to adopt manure resource utilization technologies, and the usage status of relevant machinery; and (4) Application of Intelligent Equipment, including the adoption status and compatibility of intelligent systems, automated feeding technology, the application status of precision feeding systems, and willingness to adopt precision feeding technology.

A total of 412 valid questionnaires were collected, with a sample efficiency of 98.1%. The sample number of dairy farms represents 2.75% of China’s total large-scale dairy operations, yet accounts for 18–22% of national raw milk production and 9.8% of the standardized dairy herd inventory, demonstrating strong representativeness. Due to the small sample size from areas around large cities (53), the samples from Beijing (8) and Tianjin (8) were reclassified into the North China production area, while Shanghai (2) and Chongqing (23) were included in the production area of Southern China based on geographical location. Among the survey samples, 287 adopted source reduction technology, 275 adopted process control technology, and 309 adopted end-of-pipe treatment technology, accounting for 69.2%, 66.3%, and 74.5% of the total samples, respectively. The distribution of samples is detailed in

Table 1. Sample distribution.

	Type	Number of Samples	Proportion (%)
Production area	North China	172	41.7
	Northeast China and Inner Mongolia	86	21.2
	Northwest China	76	18.3
	Southern China	78	18.9
Scale (heads)	Small (100–1000)	197	47.4
	Medium (1001–3000)	138	33.3
	Large (above 3000)	80	19.3
Types of technologies for MSR utilization	Source reduction	287	69.2
	Process control	275	66.3
	End-of-pipe treatment	309	74.5

Data source: survey data collation.

.

3.2. Methods

This study examines the impact of perceived value and environmental regulation on the adoption intensity of MSR utilization technology. The adoption intensity of MSR utilization technology serves as the dependent variable. The linear regression model is as follows:

$$Y={\alpha }_0+{\alpha }_{1}VP+{\alpha }_{2}ER+{\alpha }_3{X}_i+{\epsilon }_i$$

In this formula, Y represents the adoption behaviour of MSR utilization technology, VP represents the value perception variable, ER represents the environmental regulation variable, X_i is the control variable, α_i is the parameter to be estimated, and ε_i is the random error term. The model is run by stata15.0.

Robustness tests are carried out using methods such as Winsorization to reduce the influence of extreme values in the data (e.g., replacing outliers).

The endogeneity problem is addressed by introducing instrumental variables (e.g., “distance between the farm and the regional Agricultural and Rural Bureau”) to account for potential biases arising from reverse causality, omitted variables, or measurement errors.

To further explore the moderating effect of environmental regulation on the relationship between value perception and the adoption intensity of MSR utilization technology, we introduce the product term of value perception and environmental regulation, showing as follows:

$$Y={\alpha }_0+{\alpha }_{1}VP+{\alpha }_{2}ER+{\alpha }_3(VP\times ER)+{\alpha }_4{X}_i+{\epsilon }_i$$

The purpose of the heterogeneity analysis is to explore whether the impact of environmental regulations and value perception on technology adoption differs across farms of different sizes. This helps to demonstrate which factors have a greater influence on technology adoption across different farm scales and provides insights for policy design. This study constructs regression models for grouped analysis based on the farm size. The key regression model is as follows:

$$Y_i={\alpha }_0+{\beta }_1\mathrm{VP}+{\beta }_2\mathrm{ER}+{\gamma }_i\mathrm{Control}\mathrm{Variables}+{ϵ}_i$$

where $Y_i$ represents the technology adoption intensity of farm $i$. $\mathrm{Control}\mathrm{variables}$ include factors such as farm scale, manager characteristics, etc. ${ϵ}_i$ is the error term.

Explained Variables: The explained variable is the adoption intensity of MSR utilization technology in scale dairy farms. According to the “Action Plan for Resource Utilization of Livestock and Poultry Manure (2017–2020)” and related studies [30,31], China’s MSR utilization technologies include source reduction (automatic sprinklers, separate sewers, eco-friendly livestock feed, sewage recycling, and water-saving drinking fountains), process control (deodorants, manure and sewage treatment bacteria, composting fermentation agents, and precision feeding), and terminal treatment (fertilizer production, biogas generation, earthworm breeding, and bedding material). Due to variations in adoption and technology types among survey subjects, this study selected the comprehensive application of these technologies as the explained variable. Considering potential dependencies and correlations in adoption decisions [32,33,34], measuring adoption based solely on binary or count data can introduce bias. Thus, this study uses an improved hierarchical analysis method (Cov-AHP) to calculate the adoption intensity of such technologies. Compared with the traditional AHP method, the advantage of Cov-AHP lies in its consistency check of expert scoring, which eliminates data with logical errors. By replacing subjective scores with average covariance, it avoids relying on expert judgment and better integrates the farm technology adoption data, reflecting the actual relationships between technologies, thereby enabling a more reasonable calculation of each farm’s technology adoption intensity. The following specific calculations are based on relevant research [35,36,37,38,39]: (1) Identify the main technologies for manure resource utilization; (2) Calculate the covariance matrix based on the adoption data of the main technologies, perform transformations, and construct a judgment matrix; (3) Replace subjective scores with average covariance to obtain the weights of each environmentally friendly technology; (4) Measure the CR ratio and perform consistency estimation for the judgment matrix (see the Appendix A for related results); and (5) Calculate the comprehensive weight of each indicator within the overall indicator system. The adoption intensity refers to the degree to which a farm adopts various manure and sewage resource (MSR) utilization technologies, such as source reduction, process control, and end-of-pipe treatments. In the literature, adoption intensity has often been measured through direct variables such as technology usage rates or the extent of application. However, the contribution of our study is the use of Cov-AHP to quantitatively measure adoption intensity across multiple technologies in a more comprehensive manner. This method captures not only the extent of adoption but also the integration of multiple technologies, showing the interconnectivity of these systems. It adds value to existing adoption theories by offering a more holistic and accurate measurement of technology adoption, addressing the limitations of previous studies that often considered adoption in isolation.

Explanatory Variables: This study examines the impact of value perception and environmental regulation on MSR utilization technology adoption, using “value perception” and “environmental regulation” as core explanatory variables. Value perception is assessed across economic, ecological, and social dimensions, each examined through source reduction, process control, and end-of-pipe treatment technologies. Economic value perception (economic-VP) includes farming efficiency and income, ecological value perception (ecological-VP) focuses on reducing harmful emissions and preventing water and soil pollution, and social value perception (social-VP) encompasses the dairy cow welfare, employee well-being, and local living environment improvements. For specific meanings and values see

Table 2. Value perception index construction.

Indicator	Item	Scale	Mean	Standard Deviation
Value Perception
Social Value Perception	Impact on reducing dairy cow disease occurrences	1 = No impact; 2 = Minor impact; 3 = Neutral; 4 = Impactful; 5 = Highly impactful	3.47	0.84
	Impact on improving employee satisfaction		4.31	0.93
	Impact on enhancing the living quality of local residents		4.08	0.99
Economic Value Perception	Impact on increasing farm income	1 = No impact; 2 = Minor impact; 3 = Neutral; 4 = Impactful; 5 = Highly impactful	3.87	0.88
	Impact on improving farm production efficiency		4.61	0.58
	Impact on increasing operational costs		4.24	0.87
Ecological Value Perception	Impact on reducing manure emissions and improving the surrounding environment	1 = No impact; 2 = Minor impact; 3 = Neutral; 4 = Impactful; 5 = Highly impactful	3.41	0.83
	Impact on improving the surrounding ecological environment		3.69	1.11
	Impact on reducing pollutant emissions and preventing water and soil contamination		4.06	0.99

.

Building on existing research [40,41], this paper employs exploratory factor analysis to measure the value perception index across different dimensions. Using microdata from farms, which can be influenced by subjective factors and potential missing variables, factor analysis is applied to extract common factors and accurately measure value perception levels. We calculate the ecological, economic, and social value perception indices, with descriptive statistics presented in

Table 3. Descriptive statistics of variables.

Variable		Meaning and Value Assignment		Frequencies	Mean	Standard Deviation
Explained Variable
Adoption Intensity of MSR Utilization Technology		Measured by Cov-AHP		-	2.215	1.145
Core Explanatory Variables
Perceived Value (VP)	Economical	Factor analysis		-	0	1
	Ecological	Factor analysis		-	0	1
	Social	Factor analysis		-	0	1
Environmental Regulation (ER)	Constraint	Number of environmental inspections conducted on farms by authorities each month		-	7.163	9.183
	Motivation	Whether the farms received government subsidies for adopting MSR utilization technologies	0 = no	0.634	-	-
			1 = yes	0.368	-	-
	Guidance	Whether the information on MSR utilization technologies came from government promotion	0 = no	0.655	-	-
			1 = yes	0.345
Control Variables
Personal characteristics	Age	Years old		-	44.764	9.063
	Years of farming	Years		-	14.457	7.271
	Environmental attitude	Degree of emphasis on environmental protection (1~5 = very little attention ~ very much attention)			4.549	0.899
	Social position	Do you hold a social position other than farm manager (e.g., alliance leaders of dairy industry, technical experts, or local administrators)?	0 = no	0.617	-	-
			1 = yes	0.383	-	-
Farm characteristics	Farming scale	Heads		-	2394.154	1654.653
	Years established	Years		-	12.107	7.377
	Plant area	100 mu		-	3.813	8.613
	Water cost	1~5 = very low~very high		-	3.135	0.923
	Farm type	0 = non-agent farm		0.401	-	-
		1 = agent farm		0.599	-	-
	Farm location	Distance from the farm to the nearest settlement in kilometres		-	16.581	6.122
Social network	Information channels	Whether the relevant information comes from other farms	0 = no	0.482	-	-
			1 = yes	0.517	-	-
Dummy variable for production area		1 = Northeastern and Inner Mongolia (control group)		0.212	2.334	1.001
		2 = North China		0.417	-	-
		3 = South China		0.189	-	-
		4 = Northwest China		0.183	-	-

. The generated factor scores are standardized (i.e., centralized + scaled). The characteristics of the standardized factor scores are as follows: mean = 0; standard deviation = 1.

Environmental regulation (ER) is measured in three dimensions: constraint, incentive, and guidance. Constraint regulation is gauged by the number of government inspections per month, incentive regulation by whether the farm receives government subsidies for MSR utilization technology, and guidance regulation by whether the technology is sourced from government promotion. Descriptive statistics for these variables are also shown in Table 3.

Control Variables: The study incorporates control variables such as the farm manager’s individual characteristics (age, farming experience, environmental attitude, and social status), farm attributes (scale, established years, land area, water cost, type, and location), and social network factors. Given the potential influence of herd mentality on farmers’ behavioural choices, information acquisition channels are also considered.

4. Results

4.1. Benchmarking Regression

This paper performs multicollinearity tests on the data prior to regression analysis., revealing a maximum Variance Inflation Factor (VIF) of 2.284, indicating no significant collinearity issues among independent variables (VIF < 5: low multicollinearity, 5 ≤ VIF ≤ 10: moderate multicollinearity, VIF > 10: severe multicollinearity) The benchmarking regression results (

Table 4. Benchmark regression results.

Variable	Model 1
Economic-VP	4.428 *** (0.307)
Ecological-VP	0.372 ** (0.129)
Social-VP	0.090 ** (0.045)
Constraint-ER	0.612 ** (0.233)
Incentive-ER	8.026 *** (0.714)
Guidance-ER	1.542 (0.571)
Economic-VP * Constraint-ER	−0.824 ** (0.420)
Ecological-VP * Constraint-ER	0.456 (0.265)
Social-VP * Constraint-ER	0.264 (0.162)
Economic-VP * Incentive-ER	0.014 ** (0.007)
Ecological-VP * Incentive-ER	0.467 (0.371)
Social-VP * Incentive-ER	−0.098 (0.256)
Economic-VP * Guidance-ER	0.093 (0.101)
Ecological-VP * Guidance-ER	0.054 * (0.032)
Social-VP * Guidance-ER	0.010 ** (0.005)
Age of farm manager	−2.918 (1.800)
Manager’s farming years	−0.245 ** (0.125)
Environmental attitude	0.514 (0.361)
Manager’s social position	0.052 ** (0.027)
Farm scale	−0.966 * (0.528)
Years established of farm	−0.187 (0.477)
Water cost	0.483 * (0.293)
Farm type	1.056 ** (0.499)
Farn location	−0.723 (0.429)
Information channels	0.110 ** (0.056)
Provincial dummy variables	YES
Constant	−5.469 (4.027)
R2	0.368

Note: *, **, and *** represent the 10%, 5%, and 1% significance levels, respectively, and the values in parentheses are standard errors.

, Model 1) demonstrate that value perception, environmental regulation, and control variables significantly influence the adoption intensity of MSR utilization technology in scale dairy farms.

Economic value perception significantly promotes the adoption intensity of manure resource utilization technology in large-scale dairy farms. Among the three types of perceived value, economic value perception has the most substantial impact, with a marginal effect of 6.461%. When other variables remain constant, for each one-unit increase in the farm operator’s economic value perception, the probability of adopting manure resource utilization technology increases by an average of 4.428 percentage points. This indicates that dairy farms primarily consider economic benefits—such as cost optimization and increased revenue—when deciding to adopt new technologies. In contrast, ecological value perception and social value perception have lower marginal effects of 0.371% and 0.109%, respectively, but still have a positive influence. For example, when farms perceive that the technology can reduce manure emissions and protect the environment, their likelihood of adopting the technology increases. In practice, dairy farms often face cost pressures, particularly in efficiently handling manure. If adopting technology can reduce costs or secure government subsidies, the likelihood of adoption increases. Therefore, promotion of manure resource utilization technologies should focus on the economic benefits of the technology while also enhancing the awareness of its ecological and social values to cultivate multi-dimensional value perceptions among farmers.

In terms of environmental regulation, environmental regulations, especially incentive-based regulations, have a significant positive effect on technology adoption. This study finds that incentive-based regulations (such as government subsidies) promote the adoption of manure resource utilization technologies, with a marginal effect of 12.18%. In comparison, restrictive and guiding regulations have a smaller impact. This finding aligns with real-world practices. Incentive measures, such as subsidies, directly alleviate the financial burden on dairy farms, making the adoption of the technology more feasible. In contrast, restrictive environmental regulations (such as fines or monitoring) may have limited effectiveness on their own due to the practical costs and technical adaptation challenges faced by farms. Therefore, governments should focus on incentive-based policies, such as technological subsidies and tax benefits, to better facilitate the promotion and application of these technologies.

Control variables such as the farm manager’s farming years, social position, farm scale, water costs, farm type, and information channels all passed the significance test (Model 1, Table 4). Information acquisition channels in particular play a positive role in influencing farm decisions, especially amidst technological adoption hesitancy. This is partly because farms often exhibit a herd mentality [42], adopting a cautious approach to new technologies and being significantly influenced by the decisions of neighbouring farms.

4.2. Endogeneity Treatment

This study acknowledges potential endogeneity issues stemming from causal relationships between value perception and dairy farms’ MRS utilization behaviours, as well as concerns such as omitted variables, selection bias, and measurement errors. To mitigate these challenges, the entropy weight method was employed to compute comprehensive indices for environmental regulation and value perception, mitigating endogeneity associated with single-dimensional indicators. Additionally, an instrumental variable was applied to further test endogeneity, by using the “distance between the farm and the regional Agricultural and Rural Bureau in which it is located level” [43]. Generally, closer proximity to the agricultural management department increases the frequency and intensity of environmental regulation perception, aligning with the relevance principle. Meanwhile, the distance to the closest agricultural management department has no direct connection with the adoption of MSR utilization technology [44], confirming its status as an exogenous variable.

Table 5. Endogeneity test results.

Variables	Model 2
	Coefficient	Standard Error
Environmental regulation	0.572 *	0.098
Value perception	0.539 ***	0.087
Control variables	Yes
Provincial dummy	Yes
R2	0.291
Stage-one regression results
Environmental regulation instrumental variable	0.662 ***	0.122
F statistics	15.23

Note: * and *** represent the 10% and 1% significance levels, respectively, and the values in parentheses are standard errors.

shows that the instrumental variable is significantly correlated with the endogenous variable at the 1% level, and the F statistic is 15.23, which is greater than 10, indicating there is no weak instrument in the regression. Both environmental regulation and value perception, after controlling for potential endogeneity bias, still have a significant and positive impact on the adoption of manure utilization technology.

4.3. Moderation Effect

The analysis of the moderating effect of environmental regulations on the relationship between economic value perception and technology adoption in dairy farms reveals significant findings (Table 4). Specifically, it was found that economic value perception is positively related to the adoption of manure resource utilization technologies, and this relationship is further strengthened when combined with different types of environmental regulations—constraint-based, incentive-based, and guidance-based regulations. The results show that incentive-based regulations have the strongest moderating effect. The interaction terms between economic value perception and incentive-type environmental regulations are statistically significant at the 1% level. This indicates that when dairy farms perceive potential economic benefits—such as government subsidies or financial support—they are more likely to adopt manure resource utilization technologies. In contrast, constraint-based regulations (such as penalties for non-compliance) also show significant moderating effects, though weaker than incentive-based regulations. This finding aligns with the practical understanding that dairy farmers, as rational actors, tend to prioritize profit maximization. When they are offered direct economic incentives (e.g., subsidies or financial rewards), their perceived economic value of manure resource utilization increases, making them more motivated to adopt the technology. Conversely, the deterrence effect of penalties associated with constraint-based regulations can also amplify economic value perception, prompting stronger adoption of the technologies.

The results also support H3, which stated that environmental regulations, particularly incentive-based regulations, moderate the relationship between economic value perception and technology adoption intensity, with stronger effects observed in larger farms. This is particularly evident when analysing the size of dairy farms and their response to environmental regulations. The moderating role of environmental regulations becomes even more crucial in these contexts, where incentive-based policies have a more pronounced effect compared with smaller farms that rely more on regulatory enforcement to compel adoption. This suggests that larger farms benefit more from the direct financial incentives provided by incentive-based regulations, while smaller farms may need a combination of financial support and enforcement-based policies to achieve similar adoption levels.

4.4. Heterogeneity Analysis

The study results show that economic value perception has a significant positive impact on the adoption of manure resource utilization technologies in dairy farms of all sizes, with a notable influence at the 5% level (

Table 6. Regression results by farm scale.

Variable	Model 3
	Small (100–1000)	Medium (1001–3000)	Large (Above 3000)
Economic-VP	0.083 (0.071)	0.045 (0.867)	0.081 ** (0.041)
Ecological-VP	0.084 * (0.051)	0.075 ** (0.038)	0.087 (0.053)
Social-VP	0.083 (0.051)	0.010 * (0.006)	0.081 (0.193)
Constraint-ER	0.011 *** (0.004)	0.107 ** (0.043)	0.029 (0.193)
Incentive-ER	0.283 (0.301)	0.188 * (0.101)	0.123 *** (0.048)
Guidance-ER	0.198 (0.124)	0.143 * (0.087)	0.108 ** (0.055)
Economic-VP * Constraint-ER	−0.850 ** (0.430)	−0.79 *(0.41)	−0.92 *(0.450)
Ecological-VP * Constraint-ER	0.491 * (0.270)	0.42(0.26)	0.380 *(0.250)
Social-VP * Constraint-ER	0.280 (0.170)	0.25 ** (0.16)	0.310 (0.180)
Economic-VP * Incentive-ER	0.0150 ** (0.010)	0.012 * (0.01)	0.017 (0.010)
Ecological-VP * Incentive-ER	0.481 (0.38)	0.420 (0.36)	0.530 * (0.410)
Social-VP * Incentive-ER	−0.100 (0.26)	0.080 * (0.250)	0.120 (0.270)
Economic-VP * Guidance-ER	0.100 * (0.100)	0.08 (0.09)	0.120 * (0.110)
Ecological-VP * Guidance-ER	0.057 * (0.030)	0.049 (0.030)	0.062 (0.040)
Social-VP * Guidance-ER	0.011 (0.010)	0.009 (0.001)	0.013 * (0.010)
Control variable	Yes	Yes	Yes
Provincial dummy	Yes	Yes	Yes
R2	0.312	0.267	0.361

Note: *, **, and *** represent the 10%, 5%, and 1% significance levels, respectively, and the values in parentheses are standard errors.

). However, ecological and social value perceptions exhibit significant scale heterogeneity, particularly with ecological value perception having no significant effect on small-scale farms. In practice, this suggests that economic considerations—such as cost-savings or profitability from adopting new technology—are universally important to dairy farms, regardless of their size. For small-scale farms, the adoption of manure resource utilization technologies might be limited by other factors, such as financial constraints and lack of infrastructure. While large- and medium-scale farms are more likely to adopt these technologies due to better financial stability and a stronger focus on economic gains, small-scale farms often struggle to justify the high initial costs of implementing such technologies.

Regarding environmental regulations, this study highlights that incentive-based regulations significantly influence technology adoption in farms of all sizes. Specifically, incentive-type regulations have a positive and significant impact on the adoption of manure resource utilization technologies at the 5% level. On the other hand, constraint-type and guidance-type regulations show scale heterogeneity, with small-scale farms being more dependent on regulatory enforcement to ensure compliance. For small-scale farms, which often use less sophisticated manure management practices and may not have sufficient resources to comply voluntarily, regulatory measures such as penalties or government supervision are necessary to enforce adherence to best practices [45]. In contrast, larger farms benefit more from incentive-based policies, such as subsidies, which align better with their ability to invest in and benefit from technology adoption. These findings suggest that regulatory strategies should be tailored to farm size to effectively encourage adoption, with smaller farms requiring stronger enforcement and larger farms benefiting from financial incentives.

This study also examines how regional differences affect the adoption of manure resource utilization technologies (

Table 7. Regression results by production area.

Variable	Model 4
	Northeast China and Inner Mongolia	North China	Southern China	Northwest China
Economic-VP	0.025 ** (0.012)	0.022 ** (0.011)	0.045 ** (0.023)	0.003 ** (0.001)
Ecological-VP	0.023 (0.136)	0.076 * (0.047)	0.156 ** (0.076)	0.204 (0.172)
Social-VP	0.003 (0.004)	0.002 (0.013)	0.027 * (0.016)	0.083 (0.017)
Constraint-ER	−0.028 (0.233)	0.019 (0.273)	0.007 (0.003)	0.021 (0.032)
Incentive-ER	0.067 ** (0.026)	0.042 *** (0.016)	0.237 *** (0.092)	0.083 ** (0.073)
Guidance-ER	0.083 ** (0.050)	0.034 (0.124)	0.004 ** (0.002)	0.011 (0.043)
Economic-VP * Constraint-ER	−0.740 * (0.380)	−0.881 **(0.442)	−0.81 ** (0.42)	−0.830 * (0.256)
Ecological-VP * Constraint-ER	0.530 (0.280)	0.451 * (0.270)	0.510 * (0.290)	0.471 (0.280) 5
Social-VP * Constraint-ER	0.220 (0.150)	0.271 * (0.172)	0.241 (0.162)	0.301 (0.181)
Economic-VP * Incentive-ER	0.011 ** (0.010)	0.014 (0.010)	0.013 (0.01)	0.016 * (0.01)
Ecological-VP * Incentive-ER	0.390 ** (0.350)	0.460 (0.371)	0.440 (0.361)	0.511 ** (0.400)
Social-VP * Incentive-ER	0.071 * (0.240)	−0.11 (0.26)	0.09 (0.25)	0.13 * (0.28)
Economic-VP * Guidance-ER	0.070 (0.081)	0.110 (0.101)	0.090 (0.091)	0.131 (0.122)
Ecological-VP * Guidance-ER	0.045 (0.03)	0.055 * (0.03)	0.052 * (0.03)	0.060 (0.04)
Social-VP * Guidance-ER	0.008 (0.00)	0.012 * (0.01)	0.010 * (0.01)	0.014 (0.01)
Control variable	Yes	Yes	Yes	Yes
Provincial dummy	Yes	Yes	Yes	Yes
R2	0.367	0.234	0.333	0.298

Note: *, **, and *** represent the 10%, 5%, and 1% significance levels, respectively, and the values in parentheses are standard errors.

). Economic value perception significantly influences adoption in all regions, while ecological value perception plays a stronger role in areas with limited land resources, such as North China and Southern regions, where farms face more pressure to manage waste effectively. In these regions, ecological concerns—such as reducing pollution and ensuring sustainable farming practices—become a critical factor in the decision to adopt new technologies.

In contrast, the impact of social value perception is most pronounced in Southern regions, where the social consequences of pollution and waste management are a concern. In these regions, farms are more likely to consider how their practices impact the community and public health, making social value perception an important driver of technology adoption [46].

4.5. Robustness Test

Given the complexity of micro-survey data, there may be inaccuracies leading to underestimations or overestimations, which could result in outliers at both ends of the sample distribution. To address potential biases due to outliers in micro-survey data, we followed established procedures [47,48] and employed the Winsorization method. This approach involved adjusting all variables by applying 5% and 95% percentile shrinkage. Values exceeding the 95th percentile and falling below the 5th percentile were replaced with these respective percentile thresholds. Regression analyses were then conducted using these adjusted values to ensure that extreme values do not disproportionately affect the results. The results, as presented in

Table A3. Robustness test of environmental regulations’ moderation effect.

Variable	Model 5	Model 6	Model 7	Model 8	Model 9	Model 10	Model 11
Economic-VP	6.217 ** (3.172)	0.621 *** (0.137)	0.145 ** (0.124)	0.311 *** (0.115)	0.215 ** (0.096)	0.220 *** (0.111)	0.108 ** (0.108)
Ecological-VP	0.409 ** (0.209)	0.118 * (0.338)	0.336 (0.243)	0.138 * (0.456)	0.291 (0.313)	0.103 * (0.388)	0.271 (0.229)
Social-VP	0.117 * (0.071)	0.051 (0.022)	0.029 (0.072)	0.093 (0.034)	0.019 (0.009)	0.009 (0.093)	0.011 (0.023)
Constraint-ER	0.689 * (0.419)	0.062 ** (0.054)	0.084 * (0.056)
Incentive-ER	12.003 *** (0.257)			0.786 ** (0.376)	0.684 ** (0.531)
Guidance-ER	0.870 (0.529)					0.946 * (0.217)	0.870 * (0.178)
Economic-VP * Constraint-ER	−0.411 (0.772)		−0.934 ** (0.239)
Ecological-VP * Constraint-ER	0.431 (0.260)		0.456 (0.265)
Social-VP * Constraint-ER	0.260 (0.161)		0.264 (0.162)
Economic-VP * Incentive-ER	0.018 (0.010)				0.014 ** (0.006)
Ecological-VP * Incentive-ER	0.371 (0.342)				0.467 (0.371)
Social-VP * Incentive-ER	0.060 (0.230)				−0.098 (0.256)
Economic-VP * Guidance-ER	0.061 (0.081)						0.093 (0.101)
Ecological-VP * Guidance-ER	0.043 (0.030)						0.054 * (0.041)
Social-VP * Guidance-ER	0.007 (0.001)						0.009 ** (0.001)
Control variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Provincial dummy	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Constant	−5.164 (3.735)	−0.167 *** (0.223)	−0.253 *** (0.137)	−2.238 *** (0.157)	−0.652 ** (0.289)	−0.236 *** (0.734)	−0.289 *** (0.689)
R2	0.228	0.314	0.298	0.287	0.283	0.323	0.34

Note: *, **, and *** represent the 10%, 5%, and 1% significance levels, respectively, and the values in parentheses are standard errors.

, show consistency with the benchmark regression results, confirming the robustness of our findings. We applied the same method to test the robustness of the environmental regulation moderation effect (Table A3), and the results again showed consistency with the benchmarking regression, confirming the robustness of the findings.

5. Discussion

This research finds that value perception plays a crucial role in the adoption of manure and sewage resource (MSR) utilization technologies in dairy farms. It highlights that economic value perception, especially in larger farms, is the strongest driver of technology adoption. This aligns with the observed trend where larger farms prioritize cost-saving and profitability in their decision-making processes. Ecological and social value perceptions are also significant, though their impact is more pronounced in smaller- and medium-scale farms. This suggests that smaller farms may be more motivated by ecological and social factors, such as reducing pollution or improving local living conditions, compared with larger farms where economic motivations dominate.

Previous studies on technology adoption in agriculture and environmental management (e.g., studies on manure management or resource utilization) have often focused on isolated factors, such as the impact of economic value or policy incentives. For example, a study by Mo et al. (2024) primarily explored economic incentives and found a strong positive correlation with adoption rates in dairy farming [22], which aligns with our finding that economic value perception plays a significant role in technology adoption for large-scale farms. However, our study extends this literature by incorporating interaction terms between value perception and environmental regulation, revealing that incentive-based policies are more effective when coupled with positive economic value perceptions. This contrasts with earlier studies, such as Si et al. (2023), which found that regulatory pressure alone could drive adoption but did not consider the moderating role of value perceptions in the adoption decision-making process [27]. Environmental regulations have a somewhat limited direct effect on the adoption of MSR utilization technologies.

Our findings align with existing research that suggests that economic incentives and government support are crucial in driving technology adoption in dairy farming, especially for larger farms. Incentive-based policies (such as subsidies) have been found to significantly enhance the adoption of environmentally friendly technologies, which is consistent with our results [43]. Furthermore, our research confirms that smaller farms are more sensitive to ecological value perceptions and constraint-based regulations, which is consistent with Zhao et al. (2022), whose research found that smaller farms are more likely to adopt technologies due to external pressures such as fines and environmental mandates [48].

One key difference between our study and prior studies is the inclusion of interaction terms in the model. The interaction between economic value perception and incentive-based regulations has not been fully explored in previous studies. Our research shows that for large-scale farms, these factors together have a synergistic effect, which significantly increases the likelihood of technology adoption. This finding is in contrast to Xu et al. (2022), where environmental regulations alone were considered to be the most significant driver of adoption [46], with limited attention to the role of economic perceptions.

6. Conclusions

This study provides empirical evidence that combining value perception and targeted policy interventions can accelerate the adoption of manure resource utilization technologies on dairy farms. The findings highlight the importance of aligning economic, ecological, and social values with policy incentives to achieve sustainable technology adoption across various farm scales.

This research contributes to the understanding of technology adoption by integrating both internal factors (such as value perceptions) and external factors (such as environmental regulations). It addresses gaps in previous studies that focused on isolated perspectives, offering a more holistic view of the factors influencing manure and sewage resource utilization (MSR) technologies on dairy farms. The study affirms that value perception—especially economic value perception—is a key driver for the adoption of MSR utilization technologies. The study also adds a nuanced understanding that ecological and social perceptions influence smaller farms, showing the heterogeneity in adoption behaviour. The study extends the existing literature by showing that environmental regulations—especially incentive-based ones (such as subsidies)—moderate the relationship between economic value perception and adoption intensity. This interaction suggests that policies need to be tailored to farm size to maximize adoption success.

This research suggests that policymakers should not only focus on incentive-based regulations (such as subsidies and financial support) but also integrate them with value perception strategies. This dual approach will better align economic incentives with ecological goals, encouraging greater adoption of MSR technologies. Large-scale farms should benefit from financial incentives (e.g., subsidies, tax incentives) that enhance their perceived economic benefits, while small-scale farms may require a combination of incentive-based and constraint-based policies (such as penalties or enforcement mechanisms) to encourage adoption, which have proven to be effective in disseminating best practices and accelerating technology adoption. This study emphasizes the importance of showcasing successful adoption through targeted training and exchange programs to encourage broader industry adoption of MSR technologies.

The limitations of this study mainly lie in two aspects. First, due to objective constraints, the data is cross-sectional and only covers large-scale dairy farms, making it difficult to comprehensively reflect the dynamic evolution of manure resource utilization technology adoption in dairy farming. Second, the sample lacks coverage of dairy farming areas around large cities, which reduces the comprehensiveness of the study. Therefore, future research should enhance the breadth and depth of sample coverage to more comprehensively and objectively reflect the state of manure resource utilization in dairy farming in China. At the same time, efforts should be made to collect panel data to explore the dynamic characteristics of technology adoption behaviour from a temporal perspective. Additionally, more in-depth tracking and investigation of typical cases should be conducted to derive generalizable insights from specific examples, and to summarize the patterns and mechanisms of manure resource utilization technologies.